YouSaid · the spoken record

Chris Miller

lines on the record
38
first
2022-11-14
most recent
2022-11-14
sittings or episodes
1
sources
podcast

Every line below is reproduced as it was said and linked to the record it came from. Nothing here is summarised or generated. Directory · Search · Corrections

  1. Well, I think if I were in Taiwan, the first thing I would be doing is advocating a much more robust defensive effort. And I think one of the key puzzles right now is while the rest of the world is looking at the Taiwan Straits and sensing dramatically more risk in Taiwan, if you look at their defense budget or what they're spending their defense budget on, it's harder to see major changes, even as the international environment gets much more complicated for them. I would first be saying to my government, why aren't we buying a lot more anti-ship missiles and a lot more anti-aircraft missiles? And we currently are. But if you're at TSMC, I think the challenge is that for 30 years you've benefited from an environment where you could work without much restriction with Chinese firms, American firms, with companies from all over the world, and not have to think hard about politics. And that era is now very clearly over. It's over for all companies, but especially for companies and the semiconductor industry.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  2. the military balance does continue to shift in China's favor and will shift in China's favor for at least the next couple of years opens up more uncertainty in my mind about whether in fact the economic costs of war will restrain China and I think it creates more of an opportunity for those in China who would say this is our window to act let's act now to have their voices heard and I think that this year 2022 has also been a good reminder that it's not always

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  3. The optimistic view is that the US and China and Taiwan are engaged in a sort of mutually assured economic destruction that mirrors the mutually assured destruction that nuclear weapons provided during the Cold War. And, you know, I think there's some truth to that in the sense that a war would be disastrous economically for all parties involved. But I also think that in the past just a couple years ago, it was clear that Chinese leaders were resolutely focused on economic growth as their primary metric for policy success in a way that after three years of zero COVID, after an attack on China's private sector tech firms by the Chinese government and by the fact that China's own government is signaling that growth is going to slow dramatically, it's less clear that economic growth is the primary priority of Chinese government. And so I think that, coupled with the fact that

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  4. China right now. So all of those are to China's advantage. All of those suggest that China's military will get much more benefit out of its existing access to the leading edge in computing than the Soviet Union was ever able to do. But it's still the case that China remains pretty far behind. And so the U.S. strategy is to hope that over the next decade, this gap between US and Chinese capabilities will widen and that this gap can then be applied to military equipment to make smarter, better equipment that can make up for the quantity gap that is opening and widening with China.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  5. That the Pentagon is hoping that it's a close parallel, but there are obviously differences between the two instances. I think one difference is that the Soviet Union didn't have a large consumer economy that had a big computing industry, whereas China does. So that's a clear difference, which is to China's benefit. Second is that during the Cold War, it was the Soviet Union which was voluntarily keeping itself partially cut off from the rest of the world's economy, not exclusively the book, but largely voluntarily, whereas China is quite active in trying to integrate internationally. And so that's also to China's advantage. And then thirdly, the US is starting this relatively late in the game, I think, if you compare to the scale of restrictions that were in place on the Soviet Union and the duration in which they were in place was far more grammatic compared to the restrictions that are on China.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  6. More planes, more tanks, more missiles, but we're going to have more advanced computing under girding ours, and that's going to sustain the military advantage. And that's the bet the U.S. government is making right now

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  7. Particular. That's already the case, and that by any sort of reasonable set of projections will continue into the foreseeable future. And so the only hope the U.S. is going to have of preventing the further deterioration of its military position is matching Chinese quantity with some sort of quality advantage. And if you ask yourself, what sort of quality edge does the U.S. plausibly have over the Chinese military? There are a couple of specific types of manufacturing like aircraft engines and submarine design where the US particularly does still have an advantage over China. But across military systems, the most durable edge that the US has is in computing power, the ability to produce more transistors and apply them to military problems with better sensors, better processing, more autonomy, better communications. And that's a question of semiconductors. And so the U.S. strategy is simple. China's going to produce more stuff.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  8. Close relationship, not in terms of competition from China, because the countries that competed directly with China for telecoms equipment were Sweden and Finland with Nokia and Ericsson, but rather proximity to China and therefore concern about growing Chinese power was the best predictor about how quickly a country would get concerned about Huawei equipment and their 5G networks. When it comes to semiconductors, it's pretty clear that companies don't want these restrictions in place. And we've seen pretty active lobbying by chip companies in the US and South Korea and Taiwan not to have restrictions. And instead, it's defense ministries and the defense department in the U.S. that's driving these restrictions. And their logic is really quite straightforward. It's that everyone knows that China is going to produce more ships, more planes, more drones, more missiles, more systems in general than the United States is in the Western Pacific and in the Taiwan Straits.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  9. Well, I think the question of Huawei's telecom gear should be examined somewhat separately from the question of chip production because they are different technologies. But I think in both of them, you see strategic concerns by far outweighing economic concerns. And the reality is that it's economically costly for countries to turn down Chinese products because Chinese products are sold abroad at cheap prices because the Chinese government subsidizes them. So it's a costly measure to turn them down, but governments have been increasingly doing so for security reasons. If you start with Huawei's telephone equipment, what you find is that the countries that are closest to China began raising concerns earliest about the security concerns. So it was really Japan and Australia that led the charge in terms of rejecting Huawei from their 5G networks. Then only later did the US start focusing on this issue and then year up after that. So there's a

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  10. Like Japan, like Taiwan. And in all of these countries, views of China's technological rise have shifted dramatically over the past couple of years, such that it's become far easier to explain in these countries, like in the US, why restraints on Chinese technological development might be in these countries' interests. And so as a result, although there's a couple of different countries that play a big role in the semiconductor supply chain in all

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  11. That it needs. And I think what's worth noting in this industry is that it's not so easy to say we're not going to buy from US suppliers, we're going to buy from foreign suppliers instead for two reasons. One is because there are some types of equipment where the US is the only supplier of the most advanced equipment. If you look at the machine tools needed to make semiconductors, for example, it's a Dutch firm that has 100% market share in the lithography space, which is one of the types of tools you need. But if you look at the deposition sector depositing ultra thin film several atoms thick of materials onto semiconductors, those are largely US firms that control that market. And so there simply aren't alternative sources of supply. But even beyond that particular question is the reality that most of the equipment, most of the chemicals, most of the designs and software needed to produce semiconductors are produced either by the US or by very close allies. Like the Netherlands.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  12. China is certainly trying to do that. They've been trying for almost a decade to race forward in semiconductor technology, but they're starting from a very low base. And they've got a long way to go. And if you look across the semiconductor supply chain, the different steps that are needed to design, produce, and assemble ships, China plays actually a comparatively small role in the supply chain relative to Taiwan or Korea, certainly relative to the United States. And today, China is still hugely reliant on the import of manufacturing tools from abroad, software from abroad, and also chip designs from abroad. So today China is far from self-sufficient. It's far from having a de-Americanized supply chain, which is the more plausible goal for China to pursue, and it remains deeply dependent on foreign technology, not only US technology, but especially US technology, to produce the

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  13. Used to be TSMC's second largest customer until the US forced TSMC to cut it off, but a wide variety of Chinese firms have been banned from producing at TSMC. And the US is using its influence over TSMC and over Taiwan to limit the ability of Chinese tech companies to access TSMC's production capabilities.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  14. That's right. That's right. And so today everyone relies on TSMC's chips. They sell chips to the US, to Japan, to Europe, to China. Anyone who buys a smartphone is likely buying at least one and in many cases multiple chips produced by TSMC. And so for every country, there's an strong interest in making sure that TSMC is able to keep producing chips that everyone needs. But it's also the case that because of the most advanced chipmaker in the world and because their volume is so high that the ability to control who they produce chips for, which companies they produce ships for, is an extraordinary lever of power. And the US has over the past five years been exerting more influence to cut TSMC off from producing chips for a number of Chinese firms. The most prominent example of this is Huawei, which

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  15. That's right. They're the biggest producer of chips today. They produce 90% or so of the world's most advanced processor chips. It's impossible to have a Apple product produced without them. Most smartphones couldn't be produced without their chips. A third or so of PC processors are produced in Taiwan. Chips needed for AI and data centers or for cell phone towers. The entire world economy would grind to a halt if for whatever reason TSMC's production was disrupted. And so in some ways they're the most important company in the world.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  16. Yeah. Yeah. So TSMC was founded by an entrepreneur named Morris Chang, who is one of the most, I think, underestimated entrepreneurs of the last 100 years. Which was only manufactured chips and manufactured ships that were designed by a range of other customers. So customer would design a chip, take it to him, he would manufacture it. And at the time, there was no such business model out there like it. But because the Taiwanese government was willing to fund him, he raised money from them and some other Taiwanese investors with the government help and was able to set out TSMC on this new business path that previously hadn't existed.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  17. DSMC builds a new facility every couple of years as they bring in new technology node online. So their newest technology is generally coming from their newest factory, but they still have a range of older facilities operating older types of technology.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  18. That are all just about as complicated as these to make. And you can't have a advanced semiconductor facility without having all these different types of machine tools, each of which is capable of moving materials or manipulating materials almost at the atomic level.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  19. These are the most precise types of equipment that humans have ever made. So if you take, for example, the extreme ultraviolet orthography machines produced by the Dutch company ASML, they have a 100% market share in the production of these machines. There's just no easy way to copy them. Just to give you some of the background as to how these machines are produced. They evolve the flattest mirrors humans have ever created. They have an explosion going on almost constantly inside of them at several hundred thousand degrees Fahrenheit. The amount of engineering and specialized knowledge that is required is just extraordinary, which is why it took 30 years to develop this machine, why each one cost $115 million, making it the most expensive machine tool in history, and why it takes multiple 747s to transport them. So this isn't something that you can straightforwardly copy. And you don't just need one machine. You need multiple different types of machine.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  20. So there are five companies basically that control the market for machine tools, three in the United States, one in the Netherlands, one in Japan. And you can't really make an advanced ship without buying tools from all five of them, which means you've got to be able to access these companies tools if you want to play a role in the chip industry. Once you've got their tools, you still got to learn how to make use of them in the most efficient way possible. And that's where the Taiwanese company, TSMC, is better than any other.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  21. Most of the key designers are still based in Silicon Valley or Texas is one of the other key centers of chip design that has been since the days of Texas instruments. And of the world's biggest chip designers, most are based in the United States. And even within the space of chip design, you can talk more specifically about designing certain types of chips. But if you want to design a chip, you need not only a design, but you also need the software that's necessary to actually undertake chip design. And because advanced chips will have billions of transistors on them, what that means is you need to buy ultra specialized software to undertake it. And so there are just three companies that dominate the industry for chip design software. So that's where you start. Get the software.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  22. Tools from Japan and the Netherlands, ultra-purified chemicals from Japan as well, and then produce that chip in Taiwan or in South Korea. And many other countries are completely absent from the supply chain process. And because the supply chain is so specialized with a small number of companies and countries, those countries that have a crucial part of the supply chain monopolized or is an oligopoly in their countries have an immense, not only economic position, but also an immense amount of geopolitical power that accrues to them. And that's why you've seen more and more countries over the past half decade or so try to weaponize their position in the supply chain by cutting off their adversaries' ability to access the types of goods or chemicals or equipment that they play a major role in producing.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  23. Yeah, I think globalization is a word that's thrown around a lot, but it applies very badly to the chip industry. Parts of the chip industry are globalized in the sense that today almost anywhere in the world, you can buy almost any type of advanced chip with a couple of small exceptions. And that's a global market that chip firms are selling to. But in other ways that word really completely inaccurately describes the industry because there are a tiny number of companies in countries that dominate the production of advanced chips. And most countries have close to zero role in the semiconductor supply chain and have, as a result, zero influence over chip companies and are completely reliant on this small number of countries to get the chips that they need. So today, if you look at how the chip industry is structured, if you want to make an advanced chip, you'll often use a design from the US as well as a lot of machine tools from the US.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  24. Intensive assembly work to much more technologically advance cutting edge chip making such that today the most advanced logic chips are processor chips, the types you'd find in your smartphone or PC are made in Taiwan and many of the most advanced memory chips are made in South Korea and that's been driven both by the successes of a small number of firms in those countries Taiwan's TSMC and South Korea Samsa It's been driven by the support of those two governments which have wanted to build up semiconductor industries and have spent a fair amount of money to support those firms and it's also been supported by a overall trend towards specialization whereas in the past companies would do all the process steps of creating a ship in house build their own machinery refine their own materials design their own ships produce their own ships as each part of

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  25. The production of microchips has always been a process that has multiple steps involved, and some of the steps are very capital intensive, but others are more labor intensive. And the assembly of chips, once you've got a chip, you assemble it into a package that's made out of plastics or ceramic before it's put into a final good. That assembly package has always been pretty labor intensive, and especially so during the early stage of the industry. And so in the early 1960s, there was already a move by Silicon Valley companies to set up assembly facilities in East Asia. The first was founded in Hong Kong in 1962 by Fairchild Semiconductor. And so there's been this deep interlinkage between Silicon Valley chipmakers and partners in East Asia for a long time. But over the past half century, firms in East Asia and South Korea, in Taiwan in particular, have been able to move from pretty low value labor.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  26. Grading the electronics on their systems so that they can bring new capabilities to bear. And if you look at how militaries use their equipment today relative to 50 years ago, what you find is that today there's an expectation of precision in a way that there wasn't in the past. Today we just take it for granted that missiles can be fired, fly for a couple hundred miles, and hit a target with pretty good accuracy. Whereas 50 years ago, that was an absolute fantasy that you could only imagine in a science fiction book or film. And that change has been all about computing power, both within the systems themselves, but also spread across the military to take information, communicate it securely, let it be processed, let decisions be made, and then let targeting information be transmitted to weapon systems and all that is a question of computing power that's possible now because of semiconductors. It was simply impossible half a century ago.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  27. Well, I think the answer is yes. If you ask yourself, what's the key difference between a fighter jet produced in 1972, 50 years ago, or a fighter jet produced today? It's not that they go a lot faster. It's not that they carry substantially more weaponry. It's not that the design of the plane is dramatically different, that there have been some changes with stealth designs. But the key difference is that there's a lot more computing power and communications, which is also a question of semiconductors, both on the plane, on the munitions on the plane, in the satellites communicating with the plane, on the ground, in all the different parts of the intelligence and surveillance apparatus providing targeting information to the plane. That's the key difference. And the same thing is true with tanks or artillery pieces or naval ships. And we often find militaries around the world keeping the planes and ships intact for longer than they expected and just regularly up.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  28. At certain times of producing chips of certain quality, but never at the level of quality enough to get engineers to really trust that they were going to work and be available when they needed to. So the Soviet defense industry tried hard not to rely on semiconductors, which meant that they entered the latter decades of the 20th century trying to design out computing power from their systems, just as the rest of the world and especially the US was trying to design in more and more computing power to take advantage of the new capabilities that chips were enabling.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  29. It's quite difficult. And it's A, difficult to know how to do it once. And then even if you figure out how to go from having a ship to learning how to produce one of them, it's a very different question to know how to produce them efficiently at scale. Because the key thing to do is not just to produce one. Even if it's for military uses, which are relatively small production volumes compared to smartphones or PCs, you still need a lot of them. And you need to know which ones work and which ones don't. So you've got to have the entire system scaled up to make chips that are actually useful that you'll trust in putting in your airplanes or your missile systems. And if you don't trust your tip industry, if you think that they produce lots of duds, if they don't know which ones work and which ones don't, if you can't trust that they're going to work for a long time and harsh conditions, you'll start to rely less on them. That's exactly what the Soviet Union did. They had a chip industry. It was capable.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  30. They did indeed, and it wasn't only spies, there were also exchange students. The Soviets had exchange students studying in places like Stanford University as early as the late 1950s when this was really cutting-edge technology. And several of these exchange students in the US brought chips home to the Soviet Union, handed them over. If the industry is racing forward at exponential growth rates, then copying something that's a couple years old means that you're far behind where the cutting edge is. And that's the problem the Soviets were never able to find a solution to.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  31. To produce at scale. And in this micro industry, the industry has been defined by massive economies of scale. The more you produce, the more you can learn about how to produce most efficiently, the more you can hone your production processes, and the more effective your balance you becomes, the more cost efficient it becomes. And so if you're a small producer, you're never going to keep up with the biggest players in the industry. That's true for companies competing with each other, but it's also true with countries competing with each other. And the Soviets never had the scale to keep up with what the rest of the world was able to do.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  32. I would say the first thing is that science and mass production are not the same things. And there are numerous instances where advanced scientists who have played a major role in invention of crucial technologies have failed to mass produce them. And indeed, one of the, I think, striking conclusions from my research, and I did dozens of interviews across the semiconductor industry, well, writing ship war, was the extent to which the science was the easier part and the manufacturing was the harder part. One of the engineers I spoke with who started her career in a PhD program in left academia to work at a semiconductor firm said, you know, science is actually pretty easy because you succeed by doing something once in a lab. Whereas in industry, we have to do something a billion times a day with almost perfect accuracy. And that's just far harder than doing something once. And that was where the Soviet Union really struggled because doing something once didn't get you anything.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  33. But it really struggled to work actually. The Soviets never really get their heads around innovating and focused too much on copying and they didn't have this scale that selling to a vast global market enabled them. And because they were only selling to a relatively small defense industrial sector rather than to millions of consumers around the world, their industry struggled to keep pace with what was happening in the West, whether in Europe, in Silicon Valley or eventually in Japan. And so they remained quite reliant on advances that were pioneered elsewhere and then only later after a delay brought to the Soviet Union and put into production.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  34. So the Chinese at the time were really quite far behind. And throughout the 1950s and 60s into the 70s, China was in the midst of a variety of internal disruptions caused by Mao Zedong's different policies greatly forward the culture revolution, all of which made any sort of technological or scientific advances basically impossible. But in the Soviet Union, it was a different case because the Soviets had extraordinarily smart physicists and indeed a number of physicists who won prizes, including Nobel Prizes, for their work on semiconductor physics, as well as a vast budget for building microelectronics for military uses. And so the Soviet Union tried very hard to build an advanced semiconductor industry. They relied both on domestic advances, but also by a vast campaign of trying to spy and steal technology from Europe, from the US, and eventually from Japan.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  35. That has been enabled by semiconductors. And so personal computers and then smartphones followed after those initial use cases and have driven the industry forward over the past couple of decades.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  36. The first chips were invented in nineteen fifty eight and fifty nine Texas instruments and Fairchild Semiconductor were two companies that basically simultaneously but separately invented semiconductors and they were first intended for either military or space uses and it wasn't until the early to mid-1960s that the first ships reached the civilian market. And although the earliest chips were for defense uses by the end of the 1960s, most chips were being sold for civilian purposes. At the time it was largely big computers that would have been used by major companies. And then only in the 1970s did the market for smaller consumer devices like pocket calculators, for example, really start to take off. And as chips have gotten more powerful since then, every decade has brought a new use case or a new major type of device.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  37. That's right. He envisioned home computers and personal portable communications devices, the type of thing we now call a smartphone. And at the time, it sounded like science fiction. But if you plotted forward the exponential growth that he was already seeing, it was eminently realizable as few people at the time understood.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT

  38. And what that means is that semiconductors have advanced in their capabilities far more rapidly than any other part of the economy. And if you think of what doubling every two years entails, nothing else in the economy comes close. Airplanes don't fly twice as fast every two years, houses don't become twice as big for the same price every two years. There's just no comparison anywhere else in the economy. But we've got that doubling and computing power on a very regular basis from 1965 up to the present, which is why the cost of computing and remembering data has fallen around a billionfold since the first chip was produced. And again, compare that to the rest of the economy. What else has fallen in price by a billionfold while maintaining the same level of quality? Nothing has. And that's why chips are in some ways kind of unique. The amount of productivity increase, the amount of new computing power.

    2022-11-14 · Intelligence Squared · When the Microchips are Down, with Chris Miller · IDENTIFIED FROM THE TRANSCRIPT